How to calculate correlation coefficient between two 3-d matrices?

Sagar

2014 5 23

2 个回答

更新时间：2017 10 10

24 次查看（30 天）

请先登录，再回答此问题。

请先登录再关注

请先登录，再回答此问题。

请先登录再关注

显示更早的评论

0 个投票

This is often required in analyzing climate data from netcdf files. Climate variables are stored in 3-d matrix (lat, lon, time). I want to find correlation coefficient between two 3-d matrices at each grid cell (i.e. correlation along time or 3rd dimension). Is there an easier way to do this, without reshaping, for geometric intuition of the grid map? Also how to get the p-value at each grid cell after calculating the correlation?

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

请先登录再关注

回答（2 个）

Adapa Bhagyaraja 2015-8-10

0 个投票

Hello Sagar, There is a question similar to this. Check answer at http://in.mathworks.com/matlabcentral/answers/15884-correlation-for-multi-dimensional-arrays

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

André Gadêlha 2017-10-10

0 个投票

Hi, Sagar.

Did you understand? I have got the same problem with netcdf analysis. How can I do it?

Best Regards,

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Sagar 2017-10-10

There is no way to do it directly. You must reshape the matrix and convert to vectors. After that you can calculate correlations in a loop.

请先登录，再进行评论。

请先登录，再回答此问题。

类别

在帮助中心和 File Exchange 中查找有关 NetCDF 的更多信息

标签

Sagar

2014-5-23

Sagar

2017-10-10

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Translated by